v0.1.4

Nerf v0.1.4

Diff since v0.1.3

v0.1.3

Nerf v0.1.3

Diff since v0.1.2

v0.1.2

Nerf v0.1.2

Diff since v0.1.1

v0.1.1

Nerf v0.1.1

Diff since v0.1.0

v0.1.0

Nerf v0.1.0

Merged pull requests:

• Manually free resources to avoid OOM (#1) (@pxl-th)
• Performance and stability improvements (#2) (@pxl-th)
• Remove Manifest.toml (#3) (@pxl-th)
• Sprinkle @inbounds (#4) (@pxl-th)
• Remove unnecessary copy ops (#5) (@pxl-th)
• Minor tweaks (#6) (@pxl-th)
• Update to KA 0.9 & remove runtime dispatches (#9) (@pxl-th)
• Use random background color during training (#11) (@pxl-th)
• Reduce trainer samples allocation (#12) (@pxl-th)
• Add basic docs (#14) (@pxl-th)
• Minor performance optimizations (#15) (@pxl-th)
• Use NerfUtils.jl (#16) (@pxl-th)

Closed issues:

• Curiosity Question: Is Support For M1 in the roadmap for this project? (#8)